Pelvic inflammatory disease (PID) affects about 8 percent of all U.S. women and 11 percent of African-American women of reproductive age. It is thought to result in frequent infertility, chronic pelvic pain, and recurrence. Almost all of our knowledge of the natura1 history of and prognosis after PID comes from a single Swedish prospective study initiated 40 years ago. We propose to further knowledge of the sequelae after PID by building on the nearly-completed PID Evaluation and Clinical Health (PEACH) Study, a multicenter, randomized clinical trial of inpatient versus outpatient antibiotic management among women with clinically suspected mild-to-moderate PID. Thus far, we have accomplished an average of over 2 years (range 0.6-3.6 years) of follow-up of this predominantly African-American, indigent cohort with a follow-up rate of 85 percent. In this proposed extension, we will continue to follow women enrolled in the trial (831 randomized) for an additional 3.5 years. At the end of that time period, we will draw blood for HLA testing and for convalescent heat shock protein-60 testing. These follow-up data allow us to accomplish three aims: 1) to better describe the rates of inability to achieve pregnancy, chronic pelvic pain, and recurrence after PID; 2) to develop and validate a clinical prediction rule(s) that identifies women at risk for these consequences; and 3) to identify whether inpatient treatment is more effective than outpatient treatment in reducing the risk of sequelae among prognostic subsets of women. With additional follow-up and the use of the extensive, already-collected, baseline data set including the results of the endometrial biopsy, cervical/vaginal swab testing, blood collection, clinical examination and patient interviews, we have sufficient power to formulate and validate multivariate clinical prediction tool(s). Analytic techniques will include ROC analyses, logistic regression, and CART techniques. We will also have the power to compare the effectiveness of inpatient versus outpatient treatment within risk subgroups using life table techniques and multivariate adjusted relative risks. The importance of this extension is that it represents a unique opportunity to examine the natural history of PID, to identify women who are at highest risk for sequelae, and to test the potential benefit of more intensive management among high-risk women. We gain great efficiency in the use of the extensive resources and excellent follow-up success within the PEACH cohort.